TickZOOM Decision. Open Source and FREE!

Discussion in 'Trading Software' started by greaterreturn, Dec 15, 2008.

Thread Status:
Not open for further replies.
  1. Interesting. That's very cool. I always loved RadarScreen until I decide to be a hands free, black box trader.

    By the way, this thread is discussing, hands free, black box trading systems primarily. Mostly TickZOOM but comparisons are fine.

    As far as historical testing.

    Does TradeStation build all back testing bars tick by tick? Or does it still simulate entries and exits of stops and limit orders using imaginary/extrapolated levels?

    Just wondering if that every changed.

    Sincerely,
    Wayne
     
    #81     Dec 18, 2008
  2. jprad

    jprad

    While you still can't run a strategy in RadarScreen your indicator can call macros to execute trades.

    Doesn't matter how the bars are made now that you can define a tick or minute sub-interval for look-inside-bar-backtesting.
     
    #82     Dec 19, 2008
  3. Fascinating, that's reasonable. Hows the performance? how long does it take to test 10,000,000 ticks (or 5 years of data).

    Wayne
     
    #83     Dec 19, 2008
  4. jprad

    jprad

    I wouldn't know nor care. None of my trading methodologies require a single intraday backtest of that duration.
     
    #84     Dec 19, 2008
  5. Fair enough. Mine do. And I want many symbols at the same time so performance is critical.

    What is the advantage of TickZOOM if someone never does back testing?

    I guess TickZOOM is also useful to people who want the black box platform to log off the PC/server and run strategies hands free.

    Otherwise, if they don't need either one, they many other platforms can serve the need better, I agree.

    Sincerely,
    Wayne
     
    #85     Dec 19, 2008
  6. I dont recall [two years ago] Tradestation being that fast for a backtest. There were problems in the "look inside the bar" thingy too, I resorted to building my own bars in arrays and testing on ticks, I do that with Ninjatrader and Openquant currently, it's the only way to know that the backtest and realtime runs are doing the same thing...
     
    #86     Dec 19, 2008
  7. Yes i think the basic design is good (splitting the modeling environment, dataserver and execution part). If you compare it to Neoticker, i need a separate license for my desktop for research, one for the machine that runs the systems and one for each optimization node. Plus you need to convince the datacenter to plug an USB-dongle into one of their blades (not a funny task). And each installation carries the full overhead of the complete package.

    As for the video: why bother with a video-clip, if you put together comprehensive documentation that would be much more informative than a clip on youtube, and needs to be done anyways.

    Regards,
    P.
     
    #87     Dec 19, 2008
  8. Agreed. Wow. You did a lot of work. Well, I hope TickZOOM is useful to you.

    FYI, to everyone, I'll definately get a video up the weekend. Also, I made diagram of how the pieces fit together.

    I'm going to take a wack at some documentation to put up on the tickzoom.org site also after pippi's valid comment about that.

    And it'll include the example strategy which I'll demonstrate during the video so you can get a closer look.

    Plus I'll try to post the HTML of the strategy statistics results on the Wiki.
    Since the stats are in HTML that shouldn't be too hard.

    Wayne



     
    #88     Dec 19, 2008
  9. Pippi, the video was for a specific purpose of demonstrating the "speed".

    Several imply it's a pip dream (oops I meant pipe dream) and impossible. I think they referred to my speed claims.

    Not only has TickZOOM leaped frogged all the rest to 100 X faster speed without memory limitations, some enhancement will make it have unlimited power with unlimited parallel processing.

    Sincerely,
    Wayne

     
    #89     Dec 19, 2008
  10. ok, ok, where is the source :cool: ?

    Cheers,
    Anton
     
    #90     Dec 19, 2008
Thread Status:
Not open for further replies.